university of groningen gold-based complexes bertrand, benoit · chapter 2 caffeine-based gold(i)...

University of Groningen

Gold-based complexesBertrand, Benoit

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite fromit. Please check the document version below.

Document VersionPublisher's PDF, also known as Version of record

Publication date:2015

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):Bertrand, B. (2015). Gold-based complexes: synthesis and evaluation as anticancer agents. [S.n.].

CopyrightOther than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of theauthor(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

Take-down policyIf you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediatelyand investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons thenumber of authors shown on this cover page is limited to 10 maximum.

Download date: 03-04-2021

https://research.rug.nl/en/publications/goldbased-complexes(46953245-aa27-4721-a6fb-888dc6d3e5ef).html

CHAPTER 2

Caffeine-based gold(I) N-heterocyclic carbenes as possible anticancer agents: synthesis and biological

properties

Based on

Inorg. Chem. 2014, 53, 2296-2303

ChemBioChem 2012, 13, 1905-1912

KEYWORDS Gold(I) N-heterocyclic carbenes; cytotoxicity; G-quadruplexes; precision cut tissue

slices, cancer.

CHAPTER 2: Caffeine-based gold(I) NHC as possible anticancer agents

65

Abstract

A new series of gold(I) N-heterocyclic carbene (NHC) complexes based on xanthine

ligands have been synthesized and characterized by mass spectrometry, NMR and X-ray

diffraction. The compounds have been tested for their antiproliferative properties in human

cancer cells and non-tumorigenic cells in vitro. The bis-carbene complex [Au(caffeine-2-

ylidene)2][BF4] appeared to be selective for human ovarian cancer cell lines, and poorly toxic

in a model of healthy organs ex vivo. To gain preliminary insights into their actual

mechanism of action, two biologically relevant in cellulo targets were studied, namely the

unconventional DNA structures G-quadruplexes, playing a key role in oncogenetic

regulation, and a pivotal enzyme of the DNA damage response (DDR) machinery named

PARP-1, involved in cancer resistance mechanism. Our results indicate that our Au(I) NHC

compound acts as an efficient and selective G-quadruplex stabilizer while being a modest

PARP-1 inhibitor (i.e., poor DDR impairing agent) and, thus, provide preliminary insights

into the molecular mechanism underlying its antiproliferative effects in cancer cells.


66

Introduction

Following the discovery of the cytotoxic properties of Cisplatin by Rosenberg at the end of

the 60’s,1 the interest for metal-based anticancer treatments increased tremendously.

However, in spite of their great success, administration of platinum compounds presents

important drawbacks such as severe side-effects and development of drug resistance, which

limit their domain of applicability.2 Therefore, many different inorganic and organometallic

compounds have been developed and evaluated for anticancer activity, including platinum,

ruthenium, iron, and gold complexes.3, 4, 5, 6, 7

The latter have been the object of intense

studies by our group among others.8, 9

In spite of their promising antiproliferative effects, the risk in developing gold

compounds for biological applications is the remarkable oxidizing character of the

gold(III)/gold(I) centers and a tendency to reduce to gold(I)/gold(0) leading to extensive and

unselective cell damage, as well as to possible compound’s inactivation in aqueous

environment. This is particularly true within the fairly reducing intracellular milieu.

Therefore, different families of organometallic gold complexes were synthesized in which

the presence of a direct carbon-gold bond greatly stabilizes the gold oxidation state and

guarantees more controlled chemical speciation in biological systems. In general, both

organometallic gold(I) and gold(III) compounds have increased stability compared to the

classical gold-based coordination complexes, allowing to design gold compounds in which

the redox properties and ligand exchange reactions can be modulated to achieve selective

activation in diseased cells.

Within this frame, in the last years, gold(I) N-heterocyclic carbenes (NHCs) have

transformed from niche compounds to some of the most popular scaffolds in medicinal

inorganic chemistry.10, 11

In fact, several studies have described the promising anticancer

activities of gold(I) NHC complexes in vitro, and in a few cases also in vivo.12

Concerning

the possible mechanisms of action, the antiproliferative effects of gold NHC compounds

have been shown to be mediated by strong antimitochondrial effects via inhibition of the

seleno-enzymes thioredoxin reductase (TrxR), involved in maintaining the redox

homeostasis of cells.13, 14, 15

Thus, for example, cytotoxic compounds based on a

benzimidazol-2-ylidene core, such as the chlorido-(1,3-dimethylbenzimidazol-2-

ylidene)gold(I) complex 1 ([AuCl (Me2BIm)], Figure 1),16

as well as on an imidazol-2-

ylidene scaffold17, 18, 19

have been reported to be efficient inhibitors of TrxR. However, it

should be noted that a general direct correlation between TrxR inhibition and cytotoxicity of

gold NHC complexes could not always be demonstrated, and this indicates that other

mechanisms besides TrxR inhibition might contribute to the overall pharmacological profile.

In view of these promising results, we designed new organometallic gold(I)

complexes with a NHC ligand based on a caffeine scaffold as possible cytotoxic agents (3-

10, Figure 1). Caffeine is a natural precursor of NHC because it possesses an imidazole ring.


67

Moreover, this natural compound and its analogues have recently drawn attention for their

possible therapeutic applications as anticancer agents.20

Figure 1: Structure of the gold(I) NHC complexes discussed in this study.

All compounds were tested in vitro against different human cancer cell lines (i.e. A2780,

A2780R, SKOV3 and A549) along with a model of non-cancer cells (i.e. HEK-293T). The

gold(I) NHC derivative 1 and its bis-carbene analogue 2 ([Au(Me2BIm)][BF4], Figure 1)16, 21

were also synthesized and tested against the aforementioned cells for comparison purposes.

To gain preliminary mechanistic insights, we screened the interactions of bis-NHC

complexes 2, 4-10 with both quadruplex- and duplex-DNA. The properties of some of these

complexes were further investigated ex vivo on precision-cut tissue slices (PCTS) of liver,

kidney and colon, to assess the compounds’ toxicity in healthy organs;22

PCTS are viable ex

vivo explants of tissue with a reproducible and well defined thickness, containing cells in

their natural environment. Notably, this technique is a FDA approved model for drug toxicity

and metabolism studies. Additionally, some were tested for the inhibition of the zinc-finger

protein poly-(adenosine diphosphate (ADP)-ribose) polymerase 1 (PARP-1), a key enzyme

in the DNA Damage Response (DDR), inspired by a recent report on the potent inhibition of

this zinc-finger enzyme by Au(I) and Au(III) compounds.23, 24

PARPs are key enzymes in the

DNA Damage Response (DDR), binding to single-stranded breaks and base-excision sites to

facilitate repair processes, and, therefore, they are essential proteins involved in cancer

resistance to chemotherapies, including Cisplatin.25

Studying both DNA interacting and DDR

pivotal enzyme inhibitory properties enabled us to gain preliminary insights into the actual

mechanism of action of these new gold(I) NHC complexes.


68

Results and Discussion

Synthesis and structural characterization

Initially, methylcaffeinium iodide and tetrafluoroborate were synthesized according

to published methods by quaternization of caffeine using Meerwein’s salt in refluxing

1,2-dichloroethane or a large excess of iodomethane in refluxing DMF.26, 27

Due to a lack of

nucleophilicity of caffeine, its quaternization indeed requires harsh reaction conditions and is

limited to methyl or benzyl substituents.26, 27, 28

To enlarge the scope of these ligands, we

considered an alternative route starting from theophylline. Thus, according to previously

reported data by Petch et al,29

theophylline was allowed to react with two equivalents of

substituted allyl or benzyl bromide in the presence of potassium carbonate in dry DMF at

room temperature (Scheme 1). The resulting N7-substituted theophyllines 6a-10a were

obtained in moderate to good yields (51-95%) by precipitation upon addition of water. N-

allyltheophylline (5a) did not precipitate in these conditions and was difficult to isolate; thus,

it was obtained by deprotonation of theophylline using NaH in dry THF and addition of allyl

bromide (34% yield). The next step consisted in the methylation of the N7-substituted

theophyllines (5a-10a) with Meerwein’s salt in refluxing 1,2-dichloroethane, affording the

corresponding theophyllinium tetrafluoroborates (5b-10b) in good yields (Scheme 1).

Scheme 1: Synthesis of the theophyllinium tetrafluoroborates 5b-10b.

The neutral compound [AuI(MC)] (3) was then synthesized adapting a procedure described

by Berners-Price et al., consisting of the deprotonation of the iodide salt of the methylated

caffeinium to afford the free carbene, which was further reacted with [Au(tht)Cl] (Scheme

2).30

In parallel, the different cationic complexes 4-10 were synthesized via the commonly

used silver carbene route by reacting caffeinium and theophyllinium tetrafluoroborates (5b-

10b) with silver oxide, and then with [Au(tht)Cl] (Scheme 2).21

All new Au(I) complexes

were characterized by 1H and

13C-NMR, and by

19F-NMR in the case of complex 10. The

1H

NMR spectra of complexes 3-10 are rather similar to those of the corresponding imidazolium

salts, the signal of the imidazolium proton (8.80 < δ < 9.49 ppm) being however obviously

absent. In the 13

C-NMR spectra, a significant shift is observed for the signal corresponding to


69

the carbenic carbon moving from δ = 139 ppm for the imidazolium salts to δ = 187 ppm for

the Au(I) carbene. It is worth noting that for all cationic complexes 4-10 the 1H-NMR spectra

display a perfect symmetry of the molecule, both carbene ligands giving rise to the same set

of signals even by decreasing the NMR acquisition temperature down to 213 K. This

isochrony can be explained by the fast rotation of the ligand around the Au-Ccarbene axis

compared to the NMR time scale. This observation is in good agreement with the work of

Berners-Price in the case of bis(1-ethyl-3-methylimidazol-2-ylidene)Au(I) complex.14

Scheme 2: Syntheses of the caffeine and theophylline-based NHC Au(I) complexes (3-10).

Suitable crystals for X-ray structure determination were grown from complexes 3, 4, 7 and

10 and their solid state structures were solved (see experimental section). The structures

reveal the typical linear two-coordinated geometry of the Au(I) cation with the angles C-Au-

I (for 3) and C-Au-C (for 4, 7 and 10) being 177.66(12)°, 176.02(16)°, 178.23(28) and

175.00(39)°, respectively (Figure 2). The main difference in the solid state structures of the

cationic gold(I) bis(heterocarbene) complexes 4, 7 and 10 lies in the relative position of the

two purine ligands. Both purine rings in complex 4 are coplanar with a dihedral angle of

2.27(16)° but are tilted from 73.37(12)° and 83.42(20)° in complexes 7 and 10, respectively.

We could not rationalize this difference as resulting from strong intra or intermolecular

interactions and we therefore suspect a low energy difference between planar and orthogonal

conformers. It is worth mentioning that a planar cationic silver complex isostructural to 4 has

been described,31

while cationic gold complexes bearing two benzyl substituted imidazole-

based NHC ligands oriented in orthogonal planes have been also reported in the literature.32


70

Figure 2: ORTEP views of compounds 3, 4, 7 and 10. Solvent molecules are omitted for

clarity.

As previously mentioned, caffeine and N-substituted theophylline react more sluggishly with

MeI than their methylimidazole or methylbenzimidazole analogues. Therefore, we expected

a weaker σ-bond donor ability of the corresponding xanthine-derived NHC ligands. Such

expectation is, at variance with the similitude in the Au-Ccarb bond lengths, confirmed in our

caffeine- and theophylline-derivatives in regard to imidazole-based NHC complexes:

2.004(4) Å for 3 (mean = 2.007 Å for imidazole-based AuI complexes, 2 structures in the

CCDC) and a mean of 2.023 Å for 4, 7 and 10 (mean = 2.021 Å for 84 imidazole-based

structures in the CCDC). This apparent discrepancy can be rationalized if we simply consider

the mesomeric forms of caffeine and of the methylated caffeine, the latter being isolobal to

the gold complex (see Scheme 3). The carbonyl group of the dioxopyrimidine ring next to

the imidazole moiety withdraws electrons through delocalization and thus decreases electron

density on the imidazole ring. Conversely, the opposite NMe group of the dioxopyrimidine

ring shows a +M effect and releases electrons toward the imidazolium ring. Consequently,

two significantly different bond lengths are observed in complex 3, 4, 7 and 10 for Ccarb-N:

1.338(3) Å and 1.369(3) Å as means over seven pairs of bonds,33

the shorter bond being

located on the side of the conjugated carbonyl group. Thus, the dioxopyrimidine fragment

inhibits the reactivity of the caffeine toward simple alkylation but may act as push-pull


71

substituent giving rise to stable carbene gold complex as attested by the Au-Ccarbene bond

lengths in complex 3, 4, 7 and 10.

Scheme 3: Mesomeric forms of caffeine (left) and methyl caffeinium (right).

With the synthetic conditions set up, we tried to couple the caffeine core to an organic

fluorophore to be able to visualize the caffeine-based carbene complexes in cellulo. We

choose a coumarin derivative which is well-known for cellular imaging.34

We thus

introduced the coumarin moiety by reacting theophylline with the commercially available 4-

bromomethyl-7-methoxycoumarin using previously described conditions. Using the same

conditions as previously, theophylline was coupled to the coumarin moiety with high yield

(11a, Scheme 4). Moreover, we also coupled 4-bromomethyl-7-methoxy coumarin with

theobromine (the isomer of theophylline where the NH is in position 1) to stick as close as

possible to the caffeine scaffold. This last reaction required harsher condition due to the

lower reactivity of the N1 position compared to the N7 (12a, Scheme 4).

Scheme 4: Synthesis of the 7-methoxy-4-methylenecoumarin- N1- and N7-subsituted

xanthines.

We then performed the quaternization of N9 of each xanthine using Meerwein’s salt.

Surprisingly, the presence of the coumarin moiety didn’t improve the solubility of the

compounds in organic solvents. This resulted in a lower yield for the reaction of 11a with

Meerwein’s salt (Scheme 5). This solubility issue was even worse for compound 12a making


72

the quaternization reaction poorly efficient and the subsequent separation of product and

unreacted starting material impossible.

Scheme 5: Quaternization of the N1- and N7-subsituted xanthine by 7-methoxy-4-methylene

coumarin.

11b was treated with silver oxide at room temperature in the darkness followed by reaction

with [AuCl(tht)]. Unfortunately, we never managed to obtain the expected product in pure

form. We thus decided to introduce the coumarin moiety via an ethane bridge. As a new

starting material, 7-hydroxycoumarin was reacted with a large excess of 1,2-dibromoethane

in DMF to obtain the coumarin 13 with a bromoethyl linker. 13 was then reacted under

microwave irradiation with both theophylline and theobromine to give the N7 and N1

substitution products 14a and 15a (Scheme 6). This reaction was adapted from a procedure

from the literature.35

Scheme 6: Substitution at the N7 and N1 positions by ethane-bridge coumarin.


73

As we noticed in the case of the methylene-brigded compound 12a, 15a appeared

not to be very soluble in common organic solvents. We next performed the quaternization of

the N9 position using Meerwein’s salt and just like in the case of 11a and 12a, we could only

obtain the xanthinium salt 14b corresponding to the theophylline derivative 14a (scheme 7).

Scheme 7: Quaternization of the ethane-bridge coumarin-xanthines.

In a final step, xanthinium salt 14b was reacted with silver oxide overnight and then with

[AuCl(tht)] for 24 h as performed for the synthesis of all previously mentioned Au(I)

complexes. Although, no more evolution was noticed, the reaction was not complete as

assessed by 1H NMR spectroscopy as the imidazolium proton is still visible at δ = 9.5 ppm.

We also noticed the presence of two other side-products. Despite all our tries, no coumarin-

labeled caffeine-based Au-NHC has been obtained. The replacement of the coumarin

fluorophore by a more lipophilic moiety such as bodipy or rhodamine as well as the increase

of the linker length should be envisaged for the synthesis of labeled caffeine-based Au-NHC

complexes.

In vitro cell viability assays

The antiproliferative properties of the Au(I) NHC complexes 1-10 (with Cisplatin

used as comparison) were assessed by monitoring their ability to inhibit cell growth using the

MTT assay in the human ovarian cancer A2780 cell line, its Cisplatin resistant variant

A2780/R, in human ovarian cancer SKOV3 cells, as well as in the human lung cancer A549

cell line. In addition, in order to evaluate the compounds’ “selectivity” for cancer compared

to healthy cells, the gold complexes were also tested in human embryonic kidney HEK-293T

cells. The IC50 values of the caffeine-based Au(I) complexes are presented in Table 1.


74

Table 1. IC50 values of Au(I) NHC complexes against various cancer cell lines and non-

cancer cells HEK-293T compared to Cisplatin after 72 h incubation at 37°Ca, b

.

IC50 (µM)

Complex A2780 A2780/R SKOV3 A549 HEK-293T

1 10.9 ± 1.0 - - 17.8 ± 4.8 10.9 ± 2.4

2 0.54 ± 0.12 - 0.75 ± 0.29 5.9 ± 2.2 0.20 ± 0.09

3 37 ± 15 49 ± 15 37.3 ± 9.8 25.4 ± 2.2 22.9 ± 6.9

4 16.2 ± 2.1 15.6 ± 2.7 62.7 ± 7.8 > 100 >100

5 26.0 ± 2.2 17.2 ± 1.7 60 ± 14 52.8 ± 5.2 42.0 ± 4.0

6 28.4 ± 4.0 25.8 ± 1.7 25.6 ± 4.5 46.7 ± 5.6 38.7 ± 8.3

7 12.4 ± 0.2 17.1 ± 0.4 21.8 ± 2.3 47.7 ± 0.6 32.5 ± 4.4

8 23.4 ± 4.0 20.7 ± 2.8 53.8 ± 4.6 90.0 ± 4.8 82 ± 13

9 21.9 ± 2.4 22.1 ± 3.2 37.6 ± 7.2 56.0 ± 7.9 84 ± 11

10 13.1 ± 2.4 17.8 ± 1.7 30.3 ± 3.4 26.1 ± 2.1 37.9 ± 2.1

cisplatin 5.2 ± 1.9 35 ± 7 13.2 ± 3.5 8.0 ± 0.5 11.0 ± 2.9

a Mean SE of at least three determinations or mean of three independent experiments

performed with quadruplicate cultures at each tested concentration. b

Solutions of the gold

complexes were prepared by diluting a freshly prepared stock solution (10-2

M in DMSO) of

the corresponding compounds in cell culture medium. The stability of the complexes in

DMSO was checked: after 20 h at room temperature no degradation and no ligand

replacement by DMSO was observed. Cisplatin stock solutions were prepared in MilliQ

water

Several conclusions can be drawn in light of the results displayed in Table 1:

- All compounds elicit moderate antiproliferative properties against the tested cancer cell

lines (IC50 values lying in the micromolar range). In particular, complex 2 displays an IC50

value lying in the submicromolar range. On the contrary, the MC ligand is completely non-

toxic in all the selected cell lines tested (IC50 > 200 M).

- The new compounds display certain selective antiproliferative properties, not being

cytotoxic for HEK-293T (neither for A459) but fairly active against the human ovarian

cancer cells A2780 and A2780R. This first series of results enthrones complex 4 as the most

promising compound, yet being less active than Cisplatin against A2780 cells (IC50 = 16.2 vs


75

5.2 M, respectively) but almost two-fold more potent against A2780/R cells (IC50 = 15.6 vs

35 M, respectively), while poorly effective against SKOV3 and A549 cancer cells (IC50 >

60 M) and with very low activity in the model of non-cancer cells (IC50 > 100 M).

Notably, neutral compound 3, although moderately cytotoxic in all the tested cell lines, is

also poorly selective as compound 2.

DNA-binding properties

In order to gain initial mechanistic insights, in collaboration with the group of Dr.

David Monchaud (ICMUB, Dijon, France), we subsequently investigated the properties of

all bis-NHC complexes, namely 2, 4-10, as G-quadruplex DNA stabilizers. G-quadruplexes

are peculiar nucleic acid architectures adopted by guanine-rich DNA and RNA sequences,

whose stability originate in the stacking of contiguous G-quartets (a planar and cyclic K+-

promoted association of four guanines in a Hoogsteen hydrogen-bonding arrangement).36

Quadruplexes are currently intensively studied since they are suspected to play important

roles in key cellular events. Quadruplex-forming DNA sequences are indeed found both in

eukaryotic telomeres37

and in promoter regions of identified oncogenes or in promotor region

of HIV-1.38, 39

Their stabilization by selective small molecules (also called G-quadruplex

ligands)40

is thus currently investigated as a mean to control key cellular events (telomere

homeostasis and chromosomal stability, as well as regulation of oncongene expression).

Moreover, increasing evidence now points towards a major role of quadruplex ligands as

DNA damaging agents;41

this observation being particularly important in light of DDR

defects of most cancer cells, in which PARP (and PARP-inhibitors) are playing a pivotal role

(vide infra).42

In other words, among the new strategies currently implemented in cancer

chemotherapy, the development of drugs targeting quadruplexes is extremely versatile and

promising.43

Within this frame, the interactions of complexes 2, 4-10 with quadruplex-DNA

were assessed via FRET-melting assays (FRET = Fluorescence Resonance Electroninc

Transfer), implemented in a competitive manner: briefly, experiments were performed with

the most classically used quadruplex-forming oligonucleotide F21T which mimics the

human telomere sequence (FAM-d[5’

G3(T2AG3)33’

]-TAMRA).44

The FRET-melting principle

is schematically presented in Figure 4A: it relies on the temperature-promoted unfolding of

an oligonucleotide doubly labeled with a FRET pair (herein fluoresceine phosphoramidate

(FAM or F) and tetramethyl-6-carboxyrhodamine (TAMRA or T)); the stability imparted by

a ligand (expressed as ΔT1/2 values, in °C), readily monitored through the modification of the

FRET phenomenom (fluorescence resonance energy transfer, Figure 3A), enables an easy

quantification of its apparent affinity for quadruplex-DNA. Herein, FRET-melting

experiments were carried out with F21T in absence or presence of an excess of the unlabeled

duplex-DNA competitor ds17 (d[5’

C2AGT2CGTAGTA2C33’

]/d[5’

G3T2ACTACGA2CTG23’

])

to assess the quadruplex-vs-duplex selectivity.


76

Figure 3: FRET-melting principle (A) and DNA binding properties of complexes 2, 4-10:

quadruplex affinity (B) and selectivity (C) evaluated via competitive FRET-melting assay

carried out on F21T (0.2 μM), Au(I) complexes (1 μM) without (black bars) or with 15 (blue

bars) or 50 equiv. (green bars) of competitor double-stranded DNA ds17). ΔT1/2 values are

mean of 2 to 4 experiments.

Results (summarized in Figure 3B,C) are interesting since they are in line with the

antiproliferative results to some extent: indeed, only four complexes exceeded the affinity

threshold (i.e. ΔT1/2 = 10°C, Figure 4B), namely 2, 4, 8 and 9 (with ΔT1/2 = 13.4, 14.0, 10.7,

and 13.4°C, respectively), and among them only complex 4 exceeded the selectivity

threshold (i.e. normalized ΔT1/2 = 50%, Figure 3C) with normalized ΔT1/2 = 89 and 69% in

presence of 15 and 50 equivalents of ds17 (i.e. an ability to retain 89 and 69% of its affinity

for F21T in the presence of 15 and 50 equivalents of ds17), respectively. These results, along

with that of antiproliferative studies, enabled us to select complexes 2 and 4 as the most

interesting prototypes for further studies, the former being highly active but unselective

(elevated antiproliferative effects on cancer cells and non-cancer cells, and DNA affinities

whatever its structure), and the latter being both active and exquisitely selective (toxic for

cancer cell lines only, high affinity and selectivity for quadruplex-DNA only).


77

We can postulate that the two unsubstituted aromatic rings from the benzimidazole

moieties found at the periphery of complex 2 make it an efficient intercalator able to slither

in between two base pairs (the major duplex-DNA binding site). Conversely, the two 1,3-

dimethyluracil moieties found at the periphery of complex 4 create a steric hindrance that

prevents intercalation between base pairs to some extent, thereby making it a weaker

classical duplex-DNA interacting compound. Given that the access of the quadruplex

binding site is less sterically demanding (being based on the ligand stacking onto the external

quartet), both complexes display roughly comparable quadruplex stabilization ability (Figure

3B). Therefore, the difference of quadruplex selectivity between complexes 2 and 4 probably

originates solely in the lower affinity for duplex structures of the latter.

To further characterize the very interesting selectivity of compound 4 for G-

quadruplex DNA over duplex DNA, we thus decided to study the interactions of 4 with four

different DNA architectures, namely duplex- and quadruplex-DNA and three- and four-way

DNA junctions. We selected three double-stranded DNA sequences: an intramolecular

(hairpin) duplex-DNA (hereafter named F-DS1-T), the shorter F-DS2-T and the mismatched

F-DS3-T. Beyond the F21T sequence (hereafter named F-GQ1-T), we also studied the

quadruplex sequences F-myc-T and F-kit2-T (hereafter named F-GQ2-T and F-GQ3-T,

respectively) which correspond to promotor areas of the oncogenes cmyc and ckit2,

respectively. We included examples of three way junctions (hereafter named F-TWJ1-3-T), a

DNA structure typical in replication forks, an event occuring during cellular division.45

The

last structures we studied were examples of four way junction (hereafter named F-FWJ1-2-T)

which are key intermediates of homologous recombination.46

Instances of each studied DNA

structures are depicted in Figure 4.

Figure 4: Examples of the studied DNA structures.

For comparison purposes, we also investigated the interactions between these DNA

structures and TMPyP4 (5,10,15,20-tetra(N-methyl-4-pyridyl)porphyrin) as the most studied

DNA ligand.47

When subjected to the above-mentioned array of FRET-melting experiments,

TMPyP4 indeed stabilizes very efficiently all DNA architectures, to different extent, with


78

ΔT1/2 values up to 15.3°C for duplex-DNA, 26.5°C for quadruplex-DNA, 21.3°C for TWJ

and 37.9°C for FWJ (see table 2). 4 was thus subjected to the above-mentioned array of

FRET-melting experiments: contrarily to what has been observed with TMPyP4, this

complex stabilizes quadruplex-DNA (with ΔT1/2 values up to 12.3°C, see table 2) with a very

high selectivity since the melting temperature of all other structures are not increased. This is

an interesting observation; we are currently investing efforts to elucidate the precise

quadruplex/4 interaction. Collectively, these results suggest that the nature-inspired design of

4 is valuable for the construction of interesting ligands that display affinity only for a single

DNA structure, herein G-quadruplex DNA.

Table 2. FRET-melting results of experiments carried out with eleven doubly labelled DNA

in absence or presence of TMPyP4 or 4.

DNA alone DNA + TMPyP4 DNA + 4

T1/2 (°C) T1/2 (°C) ΔT1/2 (°C) T1/2 (°C) ΔT1/2 (°C)

F-DS1-T 63.7 73.2 9.5 61.9 0 (-1.8)

F-DS2-T 61.4 73.2 11.8 60.1 0 (-1.3)

F-DS3-T 54.1 69.4 15.3 53.0 0 (-1.1)

F-GQ1-T 52.4 78.9 26.5 64.7 12.3

F-GQ2-T 63.9 83.2 19.3 69.7 5.8

F-GQ3-T 61.2 82.4 21.2 70.9 9.7

F-TWJ1-T 51.4 65.9 14.5 50.1 0 (-1.3)

F-TWJ2-T 55.7 68.7 13.0 54.3 0 (-1.4)

F-TWJ3-T 47.9 69.2 21.3 46.0 0 (-1.9)

F-FWJ1-T 45.8 83.7 37.9 44.3 0 (-1.5)

F-FWJ2-T 51.3 58.1 6.8 49.6 0 (-1.7)

PARP -1 activity assay

Inspired by recent results indicating that some cytotoxic gold compounds are

efficient inhibitors of the zinc-finger protein PARP-1, we tested complexes 2-4 on this

purified human enzyme. PARP-inhibitors are currently highly investigated for their selective

cytotoxicity properties: PARP inhibitors are indeed poorly toxic to normal cells but are

highly active against homologous recombination (HR)-defective cells, notably BRCA-

defective breast and ovarian cancers. PARP inhibitors can thus be considered as DDR


79

inhibitors,48

which can be used in combination with classical DNA damaging agents for

optimizing the therapeutic outcome. In other words, PARP inhibiting properties of complex

4 were investigated to decipher if it can be used as a double-edged sword, acting both a DNA

damaging drug (quadruplex interaction) and DDR impairing agent (PARP inhibition).

PARP-1 inhibition was indeed observed with all compounds 2 (IC50 = 1.20 ± 0.40 M), 3

(IC50 = 0.45 ± 0.10 M) and 4 (IC50 = 0.96 ± 0.13 M), albeit in a lower range than other

cytotoxic gold(III) and gold(I) complexes (showing IC50 values in the low nanomolar

range).24

These results are not surprising per se, since the PARP-1 inhibition by Au(I)

complexes is likely to occur via a direct gold binding to the zinc finger domain of the protein

after exchange of the carbene and iodido ligands. Since such ligands are less prone to

undergo ligand exchange reactions than for example chlorido and N-donor ligands, they

impede the metal center (gold) to covalently bind to the zinc finger domain. Altogether, these

results are interesting since they indicate that the highly selective antiproliferative properties

of complex 4 probably originate in a mechanisms relying rather on the non-covalent

interaction with peculiar DNA structures than on the covalent inhibition of DDR-related

enzymes. Nonetheless, they also clearly highlight that massive efforts have now to be

invested to decipher precisely the actual mechanism of action of gold(I) NHC complexes.

Ex vivo toxicity studies

Afterwards, complexes 2 and 4 were tested for their possible toxic effects in healthy

organs ex vivo in rat liver, colon and kidney tissues using the PCTS technology.22

Thus, tissue

slices were incubated with different concentrations of each gold complex, and after a certain

incubation time (24 h for liver and kidney, 5 h for colon slices, respectively) the viability of

the tissues was determined measuring its ATP content. The obtained results are presented in

Figure 5.


80

Figure 5: Viability of rat liver, colon and kidney PCTS after treatment with 2 (A) and 4 (B).

The obtained values were statistically analyzed using a t-test comparing treated samples with

controls.

As it can be observed, while 2 is highly cytotoxic already at 10 M concentration in liver,

kidney and colon slices (Figure 5A), complex 4 shows some toxicity only at 100 M and 24

h incubation in liver, as well as in kidney slices (Figure 5B).The effects of 4 on the precision

cut kidney slices (PCKS) were also assessed by histomorphology. The data are shown in

Figure 6 and confirm the findings made using the ATP data While the kidney slices were

only slightly affected by the compound up to 50 M during 24 h incubation in comparison to

the controls, PCKS incubated with 100 M 4 show necrosis indicated by the loss of nuclei,

particularly in the cortex area. The highest concentration also causes loss of normal

histologic architecture, as well as swelling and vacuolation of the tubular epithelium.

Notably, even at the highest concentration, the glomerulus is not affected, as previously

reported in the case of Cisplatin in a similar ex vivo model.49

Overall, although the ex vivo

data cannot be directly compared to the in vitro antiproliferative activities, the results

obtained in PCTS are in line with those observed for the same compounds in HEK-293T

cells, i.e. compound 2 being markedly more toxic than 4 in both cases..


81

cortex Medulla C

on

trol

25 µ

M

50 µ

M

10

0 µ

M

Figure 6: Histomorphology of PCKS (3 mg) treated with different concentrations of

compound 4 for 24 h. Slices were stained with hematoxylin-eosin. The original

magnification is 20x.


82

Conclusions

This study was initiated in the frame of ongoing studies in our laboratories aiming at

developing new organometallic gold compounds as anticancer agents. In fact, as

demonstrated by numerous studies, regulating the reactivity and redox chemistry of gold

compounds via the optimization of an appropriate organometallic scaffold may constitute a

strategy to achieve selectivity for cancer tissues, a feature that is often lacking with other

types of coordination gold complexes. Overall, the possibility of “fine-tuning” the stability of

organometallic gold complexes while maintaining their biological activity and decreasing

their side-effects is extremely attractive in the field of metallodrug development. Moreover,

the combination of very stable organometallic gold moieties with bioligand

functionalization, such as caffeine-type ligands, holds also great promise for further

investigation.50

In this context, we reported here on a series of new caffeine-based gold(I) NHC

complexes that have been synthesized and tested for their antiproliferative activities in

different cancer and non-tumorigenic cell lines in vitro in comparison to 1,3-

dimethylbenzimidazol-2-ylidene derivatives. The bis-carbene caffeine-based complex

[Au(MC)2][BF4] 4 has shown interesting anticancer properties in vitro against the human

ovarian cancer cell line A2780 and its Cisplatin resistant variant A2780/R, while appeared to

be poorly toxic in non-cancer HEK-293 cells in vitro, as well as in healthy tissues ex vivo.

The latter result is of particular importance for the future development of new metallodrugs

with reduced side-effects.

Interestingly, complex 4 has also been proved to be an efficient and selective

quadruplex interacting agent. To date, a number of G-quadruplex stabilizing small molecules

have been synthesized but often lack of selectivity when incubated with duplex DNA.51

Among them, metal complexes occupy an increasingly important role,52

but only very few of

the studies reported so far have been devoted to the study of Au complexes.53, 54

Most

importantly, the gold compounds investigated so far are mainly gold(III) coordination

complexes. Thus, the results presented herein further bolster the interest of gold compounds

in the quest for valuable G-quadruplex ligands, notably in light of the enticing quadruplex-

selectivity of the bis-carbene complex 4. Notably, the cationic derivative 2 efficiently

stabilizes quadruplex-DNA albeit with a very poor selectivity, which may account for its

high but indiscriminate antiproliferative properties and enhanced toxicity ex vivo.

This series of results is also interesting because it provides some valuable – yet

preliminary – insights into the mechanism that may underlie the observed antiproliferative

behavior of this class of compounds: while the covalent linking of PARP-1 enzyme seems

rather unlikely, the non-covalent interaction of the intact gold complexes (that is, without

any loss of carbenic ligand) with higher-order nucleic acids structures appears efficient. Even

more interesting is the parallel that can be drawn between the quadruplex and cytotoxicity

selectivity, although further studies are necessary to confirm direct DNA binding in cells.


83

In light of very recent results,55, 56

we are currently assessing the ability of these gold

complexes to recognize other quadruplexes, notably RNA quadruplexes,57

aiming at

combining these investigations with cellular uptake and biodistribution studies in order to

better fathom the molecular basis of their intracellular action.

Furthermore, the insertion of substituents in the N7 position did not lead to a

substantial improvement of the biological effects in terms of cytotoxicity and G-

quadruplexes selectivity; however, further chemical design is ongoing in our labs aimed at

expanding this family of compounds using theobromine instead of theophylline as a NHC

building block which allows for inserting functional groups in the back of the carbene ligand

(N1 position) while keeping the two methyl substituents in N7 and N9 positions.58

ABBREVIATIONS

DDR: DNA damage response; Me2Bim: 1,3-dimethylbenzimidazol-2-ylidene; MC:

methylated caffeine-2-ylidene; NHC: N-heterocyclic carbenes; PARP-1: poly-(adenosine

diphosphate (ADP)-ribose) polymerase 1; FRET: Fluorescence resonance energy transfer;

PCTS: Precision cut tissue slices; TrxR: thioredoxin reductase.


84

Experimental section

General Remarks

All chemical reactions were carried out under an atmosphere of purified argon using

Schlenk techniques. Solvents were dried and distilled under argon before use. The precursor

Au(tht)Cl59

and the imidazole- and benzimidazole-based complexes16, 60

have been

synthesized according to literature procedures. All other reagents were commercially

available and used as received. All the analyses were performed at the “Plateforme

d’Analyses Chimiques et de Synthèse Moléculaire de l’Université de Bourgogne”. The

identity and purity (≥ 95%) of the complexes were unambiguously established using high-

resolution mass spectrometry and NMR. Exact mass of the synthesized complexes were

obtained on a Thermo LTQ Orbitrap XL. 1H- (300.13, 500.13 or 600.23 MHz),

13C- (125.77

or 150.90 MHz) and 19

F- (282.38 MHz) NMR spectra were recorded on Bruker 300 Avance

III, 500 Avance III or 600 Avance II spectrometers. Chemical shifts are quoted in ppm (δ)

relative to TMS (1H and

13C) and CFCl3 (

19F), using the residual protonated solvent (

1H) or

the deuterated solvent (13

C) as internal standards, or CFCl3 as an external standard (19

F).

Infrared spectra were recorded on a Bruker Vector 22 FT-IR spectrophotometer (Golden

Gate ATR). X-ray diffraction data for 3, 4, 7 and 10 were collected on a Bruker Nonius

Kappa CCD APEX II at 115 K. Microwave reactions were carried on in an Anton Paar

Monowave 300 apparatus.

7-allyl-1,3-dimethylxanthine (5a)

A two-neck round-bottom flask is charged with NaH (276 mg, 11.5 mmol) in suspension into

freshly distilled THF (40 mL). Theophylline (1 eq., 2.08 g, 11.5 mmol) is added by portion

to the pure NaH suspension at room temperature. After the end of the addition, the reaction

mixture is refluxed for 4 h. Allyl bromide (1 eq., 1.0 mL, 11.5 mmol) is then added to the

refluxing mixture which is reacted overnight. After the mixture is cooled down to room

temperature, the pale yellow solution is filtered off and THF partially evaporated. After

addition of pentane (100 mL), the product precipitates as a white solid which can be isolated

after filtration and washing with pentane (870 mg, 34 % yield). 1H NMR (CDCl3, 300.13

MHz): 3.40 (s, 3 H, N-Me), 3.59 (s, 3 H, N-Me), 4.94 (dt, 2 H,3JH-H = 5.7 Hz,

4JH-H = 1.2 Hz,

N-CH2), 5.23 (ddt, 2JH-H = 0.9 Hz,

3JH-H = 16.8 Hz,

4JH-H = 1.2 Hz, 1 H, CH2-allyl), 5.31 (ddt,

2JH-H = 0.9 Hz,

3JH-H = 10.2 Hz,

4JH-H = 1.2 Hz, 1 H, CH2-allyl), 6.05 (m, 1 H, CHallyl), 7.26 (s,

1 H, N-CH=N). 13

C{1H} NMR (CDCl3, 75.48 MHz): 28.0 (N-CH3), 29.8 (N-CH3), 48.8 (N-

CH2), 106.9 (=C-C(O)), 119.4 (CH2-allyl), 132.1 (CHallyl), 140.7 (N-CH=N), 148.8 (N-C=),

151.7 (C(O)), 155.2 (C(O)). ESI-MS (DMSO-MeOH), positive mode exact mass for

[C10H12N4O2]+ (221.10330): measured m/z 221.10220 [M+H]

+. FT-IR (ATR, cm

-1): 3108,

2951, 1698, 1654, 1601, 1542, 1457, 1419, 1369.


85

General procedure for the synthesis of compounds 6a-11a:

A round-bottom flask is charged with theophylline and K2CO3 (2 eq.) in suspension into dry

DMF. Allyl- or benzylbromide (2 eq.) is added to the reaction mixture. The obtained white

suspension is stirred at room temperature overnight. Water (2 VDMF) is then added and the

reaction mixture is cooled down in ice for 2 h. The formed white precipitate is collected by

filtration, washed with 8 mL of water and reprecipitated in a dichloromethane/ether mixture.

After drying under vacuum, the pure product is obtained as a white powder.

(E)-7-(but-2-enyl)-1,3-dimethylxanthine (6a)

Theophylline (500 mg, 2.78 mmol), K2CO3 (767 mg, 5.56 mmol), crotyl bromide (0.6 mL,

5.56 mmol) and dry DMF (8 mL). Product: 334 mg (51 % yield). 1H NMR (CDCl3, 300.13

MHz): 1.72 (dd, 3JH-H = 5.1 Hz,

4JH-H = 1.2 Hz, 3 H, Mecrotyl), 3.40 (s, 3 H, N-Me), 3.58 (s, 3

H, N-Me), 4.85 (d, 3JH-H = 6.3 Hz, 2 H, N-CH2), 5.64-5.84 (m, 2 H, 2 CHcrotyl), 7.54 (s, 1 H,

N-CH=N). 13

C{1H} NMR (CDCl3, 75.48 MHz): 17.7 (CH3-crotyl), 28.0 (N-CH3), 29.7 (N-

CH3), 48.6 (N-CH2), 106.9 (=C-C(O)), 124.8 (CHcrotyl), 131.9 (CHcrotyl), 140.4 (N-CH=N),

148.8 (N-C=), 151.7 (C(O)), 155.2 (C(O)). ESI-MS (DMSO-MeOH), positive mode exact

mass for [C11H15N4O2]+ (235.11895): measured m/z 235.11790 [M+H]

+. FT-IR (ATR, cm

-1):

3117, 2947, 1694, 1648, 1598, 1540, 1479, 1433, 1403, 1372.

7-benzyl-1,3-dimethylxanthine (7a)

Theophylline (600 mg, 3.33 mmol), K2CO3 (920 mg, 6.66 mmol), benzyl bromide (0.81 mL,

6.66 mmol) and dry DMF (10 mL). Product: 832 mg (93 % yield).1H NMR (CDCl3, 300.13

MHz): 3.40 (s, 3 H, N-Me), 3.58 (s, 3 H, N-Me), 5.50 (s, 2 H, N-CH2), 7.34 (m, 5 H, Ph),

7.55 (s, 1 H, N-CH=N). 13

C{1H} NMR (CDCl3, 75.48 MHz): 27.0 (N-CH3), 28.7 (N-CH3),

49.3 (N-CH2), 106.0 (=C-C(O)), 127.0 (2 CHPh), 127.6 (CHPh-para), 128.1 (2 CHPh), 134.3

(CPh-ipso), 139.8 (N-CH=N), 147.9 (N-C=), 150.7 (C(O)), 154.3 (C(O)). FT-IR (ATR, cm-1

):

3106, 2948, 1699, 1652, 1544, 1472, 1447, 1403, 1368.

1,3-dimethyl-7-(4-nitrobenzyl)xanthine (8a)

Theophylline (2.0 g, 11.1 mmol), K2CO3 (3.07 g, 22.2 mmol), 4-nitrobenzyl bromide (4.80 g,

22.2 mmol) and dry DMF (30 mL). Product: 3.309 g (95 % yield). 1H NMR (DMSO-D6,

300.13 MHz): 3.18 (s, 3 H, N-Me), 3.44 (s, 3 H, N-Me), 5.64 (s, 2 H, N-CH2), 7.53 (d, 3JH-H

= 8.7 Hz, 2 H, 2 CHPh), 8.20 (d, 3JH-H = 8.7 Hz, 2 H, 2 CHPh), 8.31 (s, 1 H, N-CH=N).

13C{

1H} NMR (DMSO-D6, 75.48 MHz): 28.0 (N-CH3), 30.0 (N-CH3), 49.0 (N-CH2), 106.3

(=C-C(O)), 124.3 (2 CHPh), 129.0 (2 CHPh), 143.3 (N-CH=N), 144.9 (CPh-ipso), 147.6 (N-C=),

149.1 (C-NO2), 151.5 (C(O)), 154.8 (C(O)). ESI-MS (DMSO-MeOH), negative mode exact


86

mass for [C14H12N5O4]- (314.08838): measured m/z 314.08889 [M-H]

-. FT-IR (ATR, cm

-1):

3121, 1700, 1649, 1599, 1543, 1514, 1486, 1441, 1403, 1378.

7-(4-(methoxycarbonyl)benzyl)-1,3-dimethylxanthine (9a)


6.66 mmol) and dry DMF (10 mL). Product: 1.005 g (92 % yield). 1H NMR (CDCl3, 300.13

MHz): 3.39 (s, 3 H, N-Me), 3.59 (s, 3 H, N-Me), 3.90 (s, 3 H, O-Me), 5.55 (s, 2 H, N-CH2),

7.35 (d, 3JH-H = 8.4 Hz, 2 H, 2 CHPh), 7.61 (s, 1 H, N-CH=N), 8.02 (d,

3JH-H = 8.4 Hz, 2 H, 2

CHPh). 13

C{1H} NMR (CDCl3, 75.48 MHz): 28.0 (N-CH3), 29.8 (N-CH3), 49.9 (N-CH2),

52.2 (O-CH3), 106.9 (=C-C(O)), 127.6 (2 CHPh), 130.4 (2 CHPh), 130.5 (C-COOMe), 140.3

(CPh-ipso), 140.9 (N-CH=N), 149.0 (N-C=), 151.6 (C(O)), 155.2 (C(O)), 166.4 (COOMe).

ESI-MS (DCM-MeOH), positive mode exact mass for [C16H16N4O4Na] (351.10638):

measured m/z 351.10476 [M+Na]+. FT-IR (ATR, cm

-1): 3109, 2955, 1713, 1694, 1646, 1543,

1481, 1429, 1374, 1277, 1232, 1192, 1101.

1,3-dimethyl-7-(4-(trifluoromethyl)benzyl)xanthine (10a)


6.66 mmol) and dry DMF (10 mL). Product: 1.009 g (90 % yield). 1H NMR (CDCl3, 300.13

MHz): 3.37 (s, 3 H, N-Me), 3.57 (s, 3 H, N-Me), 5.54 (s, 2 H, N-CH2), 7.40 (d, 3JH-H = 8.1

Hz, 2 H, 2 CHPh), 7.60 (s+d, 3JH-H = 8.1 Hz, 3 H, N-CH=N + 2 CHPh).

13C{

1H} NMR

(CDCl3, 75.48 MHz): 28.0 (N-CH3), 29.8 (N-CH3), 49.6 (N-CH2), 106.9 (=C-C(O)), 126.1

(q, 2JC-F = 3.8 Hz, C-CF3), 128.0 (4 CHPh), 139.4 (CPh-ipso), 140.9 (N-CH=N), 149.0 (N-C=), ,

151.6 (C(O)), 155.2 (C(O)). 19

F{1H} NMR (CDCl3, 282.38 Hz): -62.8 (s, 3 F, CF3). ESI-MS

(DMSO-MeOH), positive mode exact mass for [C15H13N4O2F3Na]+ (361.08828): measured

m/z 361.08802 [M+Na]+. FT-IR (ATR, cm

-1): 3092, 1701, 1651, 1547, 1472, 1457, 1323,

1165, 1108.

Scheme 8: Labeling of the position of the protons on the coumarin scaffold.


87

7-((7-methoxy-2-oxo-2H-chromen-4-yl)methyl)-1,3-dimethylxanthine (11a)

Theophylline (400 mg, 2.22 mmol), Na2CO3 (471 mg, 4.44 mmol), 4-(bromomethyl)-7-

methoxy-2H-chromen-2-one (717 mg, 2.67 mmol) and dry DMF (10 mL). Product: 799 mg

(98% yield). 1H NMR (CDCl3, 300.13 MHz): 3.37 (s, 3 H, N-Me), 3.63 (s, 3 H, N-Me), 3.89

(s, 3 H, O-Me), 5.58 (t, 4JH-H = 1.2 Hz, 1 H, CH

A), 5.68 (d,

4JH-H = 1.2 Hz, 2 H, N-CH2), 6.85

(d, 4JH-H = 2.4 Hz, 1 H, CH

B), 6.90 (dd,

3JH-H = 9.0 Hz,

4JH-H = 2.4 Hz, 1 H, CH

C), 7.50 (d,

3JH-H = 9.0 Hz, 1 H, CH

D), 7.65 (s, 1 H, N-CH=N).

13C{

1H} NMR (CDCl3, 75.48 MHz): 27.9

(N-CH3), 29.9 (N-CH3), 47.0 (N-CH2), 55.9 (O-CH3), 101.4 (s, CHB), 106.8 (s, =C-C(O)),

109.7 (s, CHA), 110.4 (s, Cquat-CH2), 113.0 (s, CH

C), 124.1 (s, CH

D), 141.5 (N-CH=N), 149.0

(N-C=), 149.8 (s, Cquat-CHD), 151.6 (C(O)xanth.), 155.2 (C(O)xanth.), 155.6 (s, Cquat-OC(O)),

160.4 (s, C(O)coum.), 163.4 (s, Cquat-OMe). ESI-MS (MeOH), positive mode exact mass for

[C18H16N4O5Na]+ (391.10129): measured m/z 391.10034 [M+Na]

+. FT- IR (ATR, cm

-1): IR

(ATR, cm-1

): 3069, 2955, 1696, 1646, 1613, 1553, 1469, 1450, 1431, 1393, 1331, 1294,

1222, 1145, 1028.

1-((7-methoxy-2-oxo-2H-chromen-4-yl)methyl)-3,7-dimethylxanthine (12a)

A round-bottom flask was charged with theobromine (2 eq., 800 mg, 4.44 mmol), Na2CO3 (4

eq., 941 mg, 8.88 mmol) and 4-(bromomethyl)-7-methoxy-2H-chromen-2-one (1 eq., 1.25 g,

2.67 mmol) in suspension into 16 mL of dry DMF. The suspension was stirred at 120°C

overnight. Water (30 mL) was then added and the reaction mixture was cooled down in ice

for 2 h. The formed yellow precipitate was collected by filtration, washed with 8 mL of

water and reprecipitated in a dichloromethane/ether mixture. After drying under vacuum, the

pure product was obtained as a yellow powder (768 mg, 47 % yield). 1H NMR (CDCl3,

300.13 MHz): 3.62 (s, 3 H, N-Me), 3.88 (s, 3 H, O-Me), 3.99 (s, 3 H, N-Me), 5.34 (d, 4JH-H =

1.5 Hz, 2 H, N-CH2), 5.81 (t, 4JH-H = 1.5 Hz, 1 H, CH

A), 6.84 (d,

4JH-H = 2.4 Hz, 1 H, CH

B),

6.90 (dd, 3JH-H = 8.7 Hz,

4JH-H = 2.4 Hz, 1 H, CH

C), 7.59 (s, 1 H, N-CH=N), 7.63 (d,

3JH-H =

8.7 Hz, 1 H, CHD).

13C{

1H} NMR (CDCl3, 150.94 MHz): 30.2 (s, N-CH3), 33.8 (s, N-CH3),

40.7 (s, N-CH2), 55.9 (s, O-CH3), 101.3 (s, CHB), 107.6 (s, =C-C(O)), 108.3 (s, CH

A), 111.7

(s, Cquat-CH2), 112.8 (s, CHC), 124.7 (s, CH

D), 142.4 (s, N-CH=N), 149.0 (s, N-C=), 149.8 (s,

Cquat-CHD), 151.3 (s, C(O)xanth.), 154.7 (s, C(O)xanth.), 155.7 (s, Cquat-OC(O)), 161.3 (s,

C(O)coum.), 163.0 (s, Cquat-OMe).

7-(2-bromoethoxy)-2H-chromen-2-one (13)

A two-neck round bottom flask was charged with 7-hydroxycoumarine (1 eq., 1.5 g, 9.3

mmol) and Na2CO3 (3 eq., 27.8 mmol, 3.0 g) in suspension in 30 mL of DMF. 1,2-

dibromoethane (20 eq., 16 mL, 185 mmol) was added and the mixture was heated at 70°C

for 3 days. After addition of water (60 mL), the product was extracted three times with


88

dichloromethane (50 mL). Organic phases were combined and washed two times with water

(30 mL). The solution was concentrated under reduced pressure and the product was

precipitated upon addition of an excess of pentane. The product was purified by column

chromatography (silica, Et2O/DCM, 7/3). The solvants were evaporated to give the pure

product (1.36 g, 50 % yield). 1H NMR (CDCl3, 300.13 MHz, 300 K): 3.66 (t,

3JH-H = 6.0 Hz,

CH2-Br), 4.35 (t, 3JH-H = 6.0 Hz, O-CH2), 6.27 (d,

3JH-H = 9.3 Hz, CH

A), 6.80 (d,

4JH-H = 2.4

Hz, CHE), 6.86 (dd,

3JH-H = 8.7 Hz,

4JH-H = 2.4 Hz, CH

D), 7.38 (d,

3JH-H = 8.7 Hz, CH

C), 7.62

(d, 3JH-H = 9.3 Hz, CH

B).

General procedure for the synthesis of compounds 14a and 15a

A microwave 10mL-tube was charged with 13 (1 eq.), Na2CO3 (2 eq.) and

theophylline/theobromine (1 eq.) in DMF. The reaction was maintained 1 hour at 100°C

under microwave irradiation (fast heating, 150W) with magnetic stirring (600 rpm). DMF

was removed under vacuum to give the crude products which was then dissolved into

dichloromethane and washed three times with water. After drying over MgSO4 and removal

of dichloromethane under vacuum, the products were obtained as off white powders.

1,3-dimethyl-7-(2-(2-oxo-2H-chromen-7-yloxy)ethyl)xanthine (14a)

13 (101 mg, 0.375 mmol), Na2CO3 (80 mg, 0.750 mmol), theophylline (68 mg, 0.375 mmol)

and DMF (5 mL). Product: 106 mg (77 % yield). 1H NMR (CDCl3, 300.13 MHz): 3.41 (s, 3

H, N-Me), 3.59 (s, 3 H, N-Me), 4.40 (t, 3JH-H = 5.1 Hz, 2 H, N-CH2), 4.74 (t,

3JH-H = 5.1 Hz,

2 H, O-CH2), 6.26 (d, 3JH-H = 9.6 Hz, 1 H, CH

E), 6.78 (m, 2 H, CH

A + CH

B), 7.35 (d,

3JH-H =

9.3 Hz, 1 H, CHC), 7.61 (d,

3JH-H = 9.6 Hz, 1 H, CH

D), 7.70 (s, 1 H, N-CH=N).

13C{

1H}

NMR (CDCl3, 75.48 MHz): 28.2 (s, N-CH3), 30.0 (s, N-CH3), 46.4 (s, N-CH2), 66.8 (s, O-

CH3), 102.0 (s, CHE), 106.7 (s, =C-C(O)), 112.5 (s, CH

D), 113.4 (s, Cquat-CH2), 114.0 (s,

CHA), 129.1 (s, CH

C), 142.3 (s, N-CH=N), 143.2 (s, CH

B), 149.3 (s, N-C=), 149.3 (s,

C(O)xanth.), 151.8 (s, C(O)xanth.), 155.9 (s, Cquat-OC(O)), 160.9 (s, C(O)coum.), 161.0 (s, Cquat-

OMe). ESI-MS (DCM-MeOH), positive mode exact mass for [C18H16N4O5Na]+ (391.10129):

measured m/z 391.10063 [M+Na]+.

3,7-dimethyl-1-(2-(2-oxo-2H-chromen-7-yloxy)ethyl)xanthine (15a)

13 (300 mg, 1.11 mmol), Na2CO3 (236 mg, 2.22 mmol), theobromine (201 mg, 1.11 mmol)

and DMF (15 mL). Product: 302 mg (74 % yield). 1H NMR (DMSO-D6, 300.13 MHz): 3.42

(s, 3 H, NMe), 3.89 (s, 3 H, NMe), 4.28 (s, 4 H, 2xCH2), 6.27 (d, 3JH-H = 9.3 Hz, CH

E), 6.93

(dd, 3JH-H = 9.3 Hz,

4JH-H = 2.4 Hz, 1 H, CH

B), 7.01 (d,

4JH-H = 2.4 Hz, 1 H, CH

A), 7.60 (d,

3JH-H = 9.3 Hz, 1 H, CH

C), 7.96 (d,

3JH-H = 9.3 Hz, 1 H, CH

D), 7.02 (s, 1 H, N-CH=N).


89

General procedure for the synthesis of compounds 5b-11b and 14b:

A Schlenk tube was charged with Meerwein’s salt (1.5 eq.) and 5a-11a in distilled

1,2-dichloroethane. The reaction was maintained under reflux for 24 h. The solvent was

partially removed under vacuum to give a white precipitate. Precipitation was completed by

adding diethylether and the solvents were eliminated by filtration. After addition of

dichloromethane and precipitation with diethylether, the formed white precipitate was dried

under vacuum to afford the pure product.

7-allyl-1,3,9-trimethylxanthinium tetrafluoroborate (5b)

Meerwein’s salt (242 mg, 1.63 mmol), 5a (200mg, 1.36 mmol) and distilled

1,2-dichloroethane (6 mL). Product: 292 mg (67 % yield). 1H NMR (DMSO-D6, 300.13

MHz): 3.27 (s, 3 H, N-Me), 3.74 (s, 3 H, NMe), 4.15 (s, 3 H, +N-Me), 5.11 (d,

3JH-H = 5.4

Hz, 2 H,+N-CH2), 5.31 (dd,

2JH-H = 0.9 Hz,

3JH-H = 17.1 Hz, 1 H, CH2-allyl), 5.38 (dd,

2JH-H =

0.9 Hz, 3JH-H = 10.5 Hz, 1 H, CH2-allyl), 6.08 (m, 1 H, CHallyl), 9.34 (s, 1 H, N-CH=N

+).

13C{

1H} NMR (DMSO-D6, 75.48 MHz): 28.9 (N-CH3), 31.8 (N-CH3), 37.5 (

+N-CH3), 50.7

(+N-CH2), 107.5 (=C-C(O)), 120.7 (CH2-allyl), 131.4 (CHallyl), 139.8 (N-CH=N

+), 140.2 (N-

C=), 150.7 (C(O)), 153.4 (C(O)). ESI-MS (DMSO-MeOH), positive mode exact mass for

[C11H15N4O2]+ (235.11895): measured m/z 235.11782 [M-BF4]

+. FT-IR (ATR, cm

-1): 3176,

3137, 3110, 2959, 1715, 1670, 1581, 1545, 1461, 1434, 1046.

(E)-7-(but-2-enyl)-1,3,9-trimethylxanthinium tetrafluoroborate (6b)

Meerwein’s salt (152 mg, 1.03 mmol), 6a (200 mg, 0.86 mmol) and distilled

1,2-dichloroethane (4 mL). Product: 282 mg (98 % yield). 1H NMR (CDCl3, 300.13 MHz):

1.74 (dt, 3JH-H = 6.6 Hz,

4JH-H = 0.8 Hz, 3 H, Mecrotyl), 3.39 (s, 3 H, N-Me), 3.79 (s, 3 H, N-

Me), 4.19 (s, 3 H, +N-Me), 5.01 (d,

3JH-H = 6.9 Hz, 2 H,

+N-CH2), 5.64-5.74 (m, 1 H, CH-

CH2), 6.02-6.13 (m, 1 H, CH-Me), 8.80 (s, 1 H, N-CH=N+).

13C{

1H} NMR (CDCl3, 75.48

MHz): 17.8 (CH3-crotyl), 28.8 (N-CH3), 31.5 (N-CH3), 37.3 (+N-CH3), 51.2 (

+N-CH2), 107.8

(=C-C(O)), 121.8 (CHcrotyl), 136.4 (CHcrotyl), 138.5 (N-CH=N+), 139.7 (N-C=), 150.3 (C(O)),

153.2 (C(O)). ESI-MS (DMSO-MeOH), positive mode exact mass for [C12H17N4O2]+

(249.13460): measured m/z 249.13341 [M-BF4]+. FT-IR (ATR, cm

-1): 3166, 3106, 2959,

1717, 1674, 1577, 1541, 1458, 1053.

7-benzyl-1,3,9-trimethylxanthinium tetrafluoroborate (7b)



MHz): 3.27 (s, 3 H, N-Me), 3.72 (s, 3 H, N-Me), 4.15 (s, 3 H, +N-Me), 5.72 (s, 2 H,

+N-


90

CH2), 7.43 (m, 5 H, Ph), 9.46 (s, 1 H, N-CH=N+).

13C{

1H} NMR (DMSO-D6, 75.48 MHz):

28.4 (N-CH3), 31.3 (N-CH3), 37.1 (+N-CH3), 51.1 (

+N-CH2), 106.9 (=C-C(O)), 128.1 (2

CHPh), 128.8 (CHPh-para), 128.9 (2 CHPh), 134.1 (CPh-ipso), 139.4 (N-CH=N+), 139.8 (N-C=),

150.1 (C(O)), 153.1 (C(O)). FT-IR (ATR, cm-1

): 3167, 3101, 1720, 1673, 1582, 1545, 1467,

1056.

1,3,9-trimethyl-7-(4-nitrobenzyl)xanthinium tetrafluoroborate (8b)



MHz): 3.23 (s, 3 H, N-Me), 3.75 (s, 3 H, N-Me), 4.18 (s, 3 H, +N-Me), 5.86 (s, 2 H,

+N-

CH2), 7.66 (d, 3JH-H = 8.7 Hz, 2 H, 2 CHPh), 8.26 (d,

3JH-H = 8.7 Hz, 2 H, 2 CHPh), 9.49 (s, 1

H, N-CH=N+).

13C{

1H} NMR (DMSO-D6, 75.48 MHz): 28.9 (N-CH3), 31.8 (N-CH3), 37.7

(+N-CH3), 51.0 (

+N-CH2), 107.5 (=C-C(O)), 124.3 (2 CHPh), 129.6 (2 CHPh), 140.4 (C-NO2),

140.5 (N-CH=N+), 141.9 (CPh-ipso), 148.0 (N-C=), 150.6 (C(O)), 153.4 (C(O)). ESI-MS

(DMSO-MeOH), positive mode exact mass for [C15H16N5O4]+ (330.11968): measured m/z

330.11833 [M-BF4]+. FT-IR (ATR, cm

-1): 3111, 3018, 1722, 1673, 1574, 1542, 1520, 1457,

1345, 1026.

7-(4-(methoxycarbonyl)benzyl)-1,3,9-trimethylxanthinium tetrafluoroborate (9b)



MHz): 3.23 (s, 3 H, N-Me), 3.73 (s, 3 H, N-Me), 3.84 (s, 3 H, O-Me), 4.15 (s, 3 H, +N-Me),

5.79 (s, 2 H, +N-CH2), 7.53 (d,

3JH-H = 8.1 Hz, 2 H, 2 CHPh), 7.97 (d,

3JH-H = 8.4 Hz, 2 H, 2

CHPh), 9.46 (s, 1 H, N-CH=N+).

13C{

1H} NMR (CDCl3, 75.48 MHz): 28.9 (N-CH3), 31.8 (N-

CH3), 37.6 (+N-CH3), 51.3 (

+N-CH2), 52.8 (O-CH3), 107.5 (=C-C(O)), 128.7 (2 CHPh), 130.0

(2 CHPh), 130.3 (C-COOMe), 139.8 (N-CH=N+), 140.3 (CPh-ipso), 140.4 (N-C=), 150.6

(C(O)), 153.5 (C(O)), 166.2 (COOMe). ESI-MS (DCM-MeOH), positive mode exact mass

for [C17H19N4O4] (343.14008): measured m/z 343.13902 [M-BF4]+. FT-IR (ATR, cm

-1):

3105.1, 1720, 1681, 1638, 1580, 1539, 1456, 1427, 1347, 1261, 1012.

1,3,9-trimethyl-7-(4-(trifluoromethyl)benzyl)xanthinium tetrafluoroborate (10b)



MHz): 3.25 (s, 3 H, N-Me), 3.74 (s, 3 H, N-Me), 4.16 (s, 3 H, +N-CH3), 5.81 (s, 2 H,

+N-

CH2), 7.64 (d, 3JH-H = 8.1 Hz, 2 H, 2 CHPh), 7.80 (d,

3JH-H = 8.1 Hz, 2 H, 2 CHPh), 9.48 (s, 1

H, N-CH=N+).

13C{

1H} NMR (DMSO-D6, 75.48 MHz): 28.9 (N-CH3), 31.8 (N-CH3), 37.6

(+N-CH3), 51.1 (

+N-CH2), 107.5 (=C-C(O)), 126.2 (q,

2JC-F = 4.5 Hz, C-CF3), 129.3 (4 CHPh),


91

139.2 (N-CH=N+), 140.3 (CPh-ipso), 140.4 (N-C=), 150.6 (C(O)), 153.5 (C(O)).

19F{

1H} NMR

(DMSO-D6, 282.38 Hz): -61.1 (s, 3 F, CF3), -148.4 (d, BF4-). ESI-MS (DMSO-MeOH),

positive mode exact mass for [C16H16N4O2F3]+ (353.12199): measured m/z 353.12103 [M-

BF4]+. FT-IR (ATR, cm

-1): 3175, 3109, 1718, 1677, 1579, 1539, 1424, 1305, 1174, 1047.

1-((7-methoxy-2-oxo-2H-chromen-4-yl)methyl)-3,7,9-trimethylxanthinium

tetrafluoroborate (11b)



MHz): 3.23 (s, 3 H, N-Me), 3.78 (s, 3 H, N-Me), 3.90 (s, 3 H, O-Me), 4.20 (s, 3 H, +N-Me),

5.88 (s, 1 H, CHA), 6.03 (s, 2 H, N-CH2), 7.07 (dd,

3JH-H = 8.7 Hz,

4JH-H = 2.4 Hz, CH

C), 7.11

(d, 4JH-H = 2.4 Hz, CH

B), 7.85 (d,

3JH-H = 8.7 Hz, CH

D), 9.40 (s, 1 H, N-CH=N

+).

13C{

1H}

NMR (DMSO-D6, 75.48 MHz): 28.4 (N-CH3), 31.4 (N-CH3), 37.3 (s, +N-CH3), 47.9 (N-

CH2), 56.1 (O-CH3), 101.2 (s, CHB), 107.1 (s, =C-C(O)), 108.5 (s, CH

A), 109.9 (s, Cquat-

CH2), 112.7 (s, CHC), 125.6 (s, CH

D), 140.2 (N-CH=N

+), 140.4 (N-C=), 149.3 (s, Cquat-CH

D),

150.2 (C(O)xanth.), 152.9 (C(O)xanth.), 154.3 (s, Cquat-OC(O)), 159.8 (s, C(O)coum.), 163.0 (s,

Cquat-OMe). ESI-MS (MeOH), positive mode exact mass for [C19H16N4O5]+ (383.13500):

measured m/z 383.13469 [M-BF4]+. FT- IR (ATR, cm

-1): 3103, 2961, 1680, 1609, 1582,

1541, 1459, 1427, 1352, 1333, 1289, 1208, 1135, 1038.

1,3,9-trimethyl-7-(2-(2-oxo-2H-chromen-7-yloxy)ethyl)xanthinium tertrafluoroborate

(14a)



MHz): 3.23 (s, 3 H, N-Me), 3.74 (s, 3 H, N-Me), 4.20 (s, 3 H, +N-Me), 4.51 (t,

3JH-H = 4.8

Hz, 2 H, O-CH2), 4.93 (t, 3JH-H = 4.8 Hz, 2 H,

+N-CH2), 6.31 (d,

3JH-H = 9.3 Hz, 1 H, CH

E),

6.97 (dd, 3JH-H = 9.3 Hz,

4JH-H = 2.4 Hz, 1 H, CH

B), 7.05 (d,

4JH-H = 2.4 Hz, 1 H, CH

A), 7.65

(d, 3JH-H = 9.3 Hz, 1 H, CH

C), 7.98 (d,

3JH-H = 9.3 Hz, 1 H, CH

D), 9.45 (s, 1 H, N-CH=N

+).

13C{

1H} NMR (DMSO-D6, 75.48 MHz): 28.4 (N-CH3), 31.3 (N-CH3), 37.0 (

+N-CH3), 47.9

(s, O-CH2), 65.6 (s, +N-CH2), 101.8 (s, CH

E), 107.1 (s, =C-C(O)), 112.5 (s, CH

D), 112.9 (s,

CHA), 129.5 (s, CH

C), 139.6 142.3(s, N-C=), 139.9 (s, N-CH=N

+), 144.1 (s, CH

B), 150.1 (s,

C(O)xanth.), 153.3 (s, C(O)xanth.), 155.1 (s, Cquat-OC(O)), 160.1 (s, C(O)coum.), 160.5 (s, Cquat-

OMe). ESI-MS (MeOH), positive mode exact mass for [C19H19N4O5]+ (383.13500):

measured m/z 383.13364 [M-BF4]+.


92

(1,3,7,9-tetramethylxanthin-8-ylidene)Au(I) iodide (3)

A Schlenk tube was charged with 1,3,7,9-tetramethylxanthinium iodide (100 mg, 0.30

mmol) dissolved in DMF (2.5 mL) and 0.5 M KHMDS solution in toluene (1.2 eq., 0.72

mL, 0.36 mmol) was added. The obtained yellow solution was stirred for 15 min at room

temperature and then transferred onto a solution of [AuCl(tht)] (1 eq., 98 mg, 0.30 mmol) in

DMF (1.5 ml). After 1 h at room temperature, the expected complex was precipitated by

addition of diethylether (10 mL) and filtrated. The obtained white precipitate was washed

with diethylether (6 mL) and dried under vacuum to afford the pure product (86.3 mg, 47 %).

Crystals suitable for X-Ray diffraction have been obtained by slow evaporation of

dichloromethane. RMN 1H (CD2Cl2, 600.23 MHz): δ 3.37 (s, 3H, C(O)N(CH3)C=C), 3.77 (s,

3H, C(O)N(CH3)C(O)), 4.14 (s, 3H, N(CH3)imidaz.), 4.23 (s, 3H, N(CH3)imidaz.). RMN 13

C{1H}

(CD2Cl2, 125.77 MHz): δ 28.9 (C( O)N(CH3)C=C), 32.3 (C(O)N(CH3)C(O)), 37.7

(N(CH3)imidaz.), 39.0 (N(CH3)imidaz.), 109.0 (C(O)Csp2

N), 139.9 (NCsp2

N), 151.1 (NC(O)C=C),

δ 153,4 (C(O)urea), 187.1 (Ccarbene). FT-IR (ATR, cm-1

): 2951, 1704, 1666, 1538, 1461, 1437,

1406, 1298.

General procedure for the synthesis of compounds 4-10:

A round-bottom flask was charged with xanthinium salt 4b-10b (1 eq.), Ag2O (0.8 eq.) and

molecular sieves 4 Å which were suspended in distillated acetonitrile (10 mL) and kept away

from light. After 6 hours of stirring at room temperature, [Au(tht)Cl] (0.5 eq.) was added

and the stirring was maintained overnight at room temperature. KI (0.5 eq.) was added to the

mixture and the solution was filtrated off through Celite to give a colorless solution. After

washing the Celite with dichloromethane, removing of 2/3 of the volume of volatiles under

vacuum and precipitation by addition of ether, a white precipitate was formed which was

collected by filtration to afford the pure product as a white powder.

Bis(1,3,7,9-tetramethylxanthin-8-ylidene)Au(I) tetrafluoroborate (4)

4b (100 mg, 0.34 mmol), Ag2O (63 mg, 0.27 mmol), Au(tht)Cl (54 mg, 0.17 mmol) and KI

(28 mg, 0.17 mmol). Product: 63 mg (53 % yield). Crystals suitable for X-ray diffraction

were grown by slow diffusion of pentane into a saturated solution in dichloromethane. RMN 1H (DMSO-D6, 300.13 MHz): δ 3.23 (s, 3H, C(O)N(CH3)C=C), 3.77 (s, 3H,

C(O)N(CH3)C(O)), 4.16 (s, 3H, N(CH3)imidaz.), 4.27 (s, 3H, N(CH3)imidaz.). RMN 13

C{1H}

(DMSO-D6, 500.13 MHz): δ 28.8 (C(O)N(CH3)C=C), 32.1 (C(O)N(CH3)C(O)), 37.5

(N(CH3)imidaz.), 39.2 (N(CH3)imidaz.), 109.2 (C(O)Csp2

N), 140.9 (NCsp2

N), 151.0 (NC(O)C=C),

153.7 (C(O)urea), 187.0 (Ccarbene). ESI-MS (DMSO/MeOH, positive mode exact mass for

C18H24AuN8O4 (613.15805): measured m/z 613.16075 [M+]. FT-IR (ATR, cm

-1): 1706, 1662,

1537, 1434, 1030.


93

Bis(7-allyl-1,3,9-trimethylxanthin-8-ylidene)Au(I) tetrafluoroborate (5)


(26 mg, 0.16 mmol). Product: 102 mg (88 % yield). 1H NMR (DMSO-D6, 300.13 MHz):

3.26 (s, 3 H, N-Me), 3.77 (s, 3 H, N-Me), 4.27 (s, 3 H, N-Me), 5.11 (dd, 2JH-H = 0.9 Hz,

3JH-H

= 17.1 Hz, 1 H, CH2-allyl), 5.20 (d, 3JH-H = 5.4 Hz, 2 H, N-CH2), 5.36 (dd,

2JH-H = 0.9 Hz,

3JH-H

= 10.5 Hz, 1 H, CH2-allyl), 6.04-6.17 (m, 1 H, CHallyl). 13

C{1H} NMR (DMSO-D6, 75.48

MHz): 28.8 (N-CH3), 32.1 (N-CH3), 51.7 (N-CH2), 108.4 (=C-C(O)), 118.2 (CHallyl), 133.9

(CH2-allyl), 141.2 (N-C=), 150.9 (C(O)), 153.3 (C(O)), 187.2 (Ccarbene). ESI-MS (DMSO-

MeOH), positive mode exact mass for [C22H28N8O4]+ (665.18935): measured m/z 665.18702

[M-BF4]+. FT-IR (ATR, cm

-1): 3091, 3027, 2964, 1707, 1664, 1576, 1539, 1522, 1459, 1347,

1028.

Bis[(E)-7-(but-2-enyl)-1,3,9-trimethylxanthin-8-ylidene]Au(I) tetrafluoroborate (6)

6b (100 mg, 0.30 mmol), 55 mg of Ag2O (55 mg, 0.24 mmol), Au(tht)Cl (48 mg, 0.15

mmol) and KI (25 mg, 0.15 mmol). Product: 60.1 mg (51 % yield). 1H NMR (CDCl3, 300.13

MHz): 1.69 (dd, 3JH-H = 6.3 Hz,

4JH-H = 1.2 Hz, 3 H, Mecrotyl), 3.40 (s, 3 H, N-Me), 3.84 (s, 3

H, N-Me), 4.36 (s, 3 H, N-Me), 5.12 (d, 3JH-H = 6.0 Hz, 2 H, N-CH2), 5.65-5.73 (m, 1 H, CH-

CH2), 5.81-5.89 (m, 1 H, CH-Me). 13

C{1H} NMR (CDCl3, 75.48 MHz): 17.7 (CH3-crotyl),

28.8 (N-CH3), 32.1 (N-CH3), 39.5 (N-CH3), 51.8 (N-CH2), 108.5 (=C-C(O)), 125.3 (CHcrotyl),

131.9 (CHcrotyl), 140.5 (N-C=), 150.6 (C(O)), 153.2 (C(O)), 187.6 (Ccarbene). ESI-MS

(DMSO-MeOH), positive mode exact mass for [C24H32N8O4]+ (693.22066): measured m/z

693.21772 [M-BF4]+. FT-IR (ATR, cm

-1): 2950, 1707, 1664, 1536, 1463, 1420, 1050.

Bis(7-benzyl-1,3,9-trimethylxanthin-8-ylidene)Au(I) tetrafluoroborate (7)


(23 mg, 0.14 mmol). Product: 81 mg (73 % yield). Crystals suitable for X-Ray diffraction

were grown by slow evaporation of dichloromethane. 1H NMR (CD3CN, 300.13 MHz): 3.28

(s, 3 H, N-Me), 3.74 (s, 3 H, N-Me), 4.14 (s, 3 H, N-Me), 5.74 (s, 2 H, N-CH2), 7.30 (s, 5 H,

Ph), 9.46. 13

C{1H} NMR (CD3CN, 75.48 MHz): 27.7 (N-CH3), 31.2 (N-CH3), 38.8 (N-CH3),

50.3 (N-CH2), 108.3 (=C-C(O)), 126.9 (2 CHPh), 128.0 (CHPh-para), 128.5 (2 CHPh), 136.0

(CPh-ipso), 140.7 (N-C=), 150.5 (C(O)), 153.2 (C(O)), 187.4 (Ccarbene). ESI-MS (DMSO-

MeOH), positive mode exact mass for [C30H32N8O4Au]+ (765.22066): measured m/z

765.21838 [M-BF4]+. FT-IR (ATR, cm

-1): 1714, 1667, 1533, 1459, 1421, 1061, 1024. Anal.

Calc. for C30H32AuN8O4BF4.2CH2Cl2: C: 37.60, H: 3.55, N: 10.96; Found: C: 38.02, H: 3.27,

N: 11.86.

Bis(1,3,9-trimethyl-7-(4-nitrobenzyl)xanthin-8-ylidene)Au(I) tetrafluoroborate (8)


94

8b (100 mg, 0.24 mmol), Ag2O (45 mg, 0.19 mmol), distilled acetonitrile (10 mL), Au(tht)Cl

(38 mg, 0.12 mmol) and KI (20 mg, 0.12 mmol). Product: 61 mg (54 % yield). 1H NMR

(DMSO-D6, 300.13 MHz): 3.20 (s, 3 H, N-Me), 3.77 (s, 3 H, N-Me), 4.23 (s, 3 H, N-Me),

5.87 (s, 2 H, N-CH2), 7.42 (d, 3JH-H = 8.7 Hz, 2 H, 2 CHPh), 8.04 (d,

3JH-H = 8.7 Hz, 2 H, 2

CHPh). 13

C{1H} NMR (DMSO-D6, 75.48 MHz): 28.8 (N-CH3), 31.1 (N-CH3), 32.2 (N-CH3),

52.0 (N-CH2), 108.6 (=C-C(O)), 124.0 (2 CHPh), 128.2 (2 CHPh), 141.5 (C-NO2), 143.9 (CPh-

ipso), 147.4 (N-C=), 150.9 (C(O)), 153.4 (C(O)), 187.5 (Ccarbene). ESI-MS (DMSO-MeOH),

positive mode exact mass for [C30H30N10O8Au]+ (855.19081): measured m/z 855.18961 [M-

BF4]+. FT-IR (ATR, cm

-1): 3090, 1711, 1669, 1576, 1540, 1519, 1459, 1346 1026.

Bis(7-(4-(methoxycarbonyl)benzyl)-1,3,9-trimethylxanthin-8-ylidene)Au(I)

tetrafluoroborate (9)

9b (150 mg, 0.35 mmol), Ag2O (65 mg, 0.28 mmol), distilled acetonitrile (10 mL), Au(tht)Cl

(56 mg, 0.17 mmol) and KI (29 mg, 0.17 mmol). Product: 129 mg (79 % yield). 1H NMR

(DMSO-D6, 500.13 MHz): 3.21 (s, 3 H, N-Me), 3.76 (s, 3 H, N-Me), 3.83 (s, 3 H, O-Me),

4.22 (s, 3 H, N-Me), 5.80 (s, 2 H, N-CH2), 7.31 (d, 3JH-H = 8.5 Hz, 2 H, 2 CHPh), 7.78 (d,

3JH-

H = 8.5 Hz, 2 H, 2 CHPh).13

C{1H} NMR (CDCl3, 75.48 MHz): 28.8 (N-CH3), 32.1 (N-CH3),

39.5 (N-CH3), 52.3 (N-CH2), 52.6 (O-CH3), 108.6 (=C-C(O)), 127.4 (2 CHPh), 129.6 (C-

COOMe), 129.8 (2 CHPh), 141.5 (CPh-ipso), 141.7 (N-C=), 150.9 (C(O)), 153.4 (C(O)), 166.1

(COOMe), 187.4 (Ccarbene). ESI-MS (DMSO-MeOH), positive mode exact mass for

[C34H36N8O8Au]+ (881.23161): measured m/z 881.22999 [M-BF4]

+. FT-IR (ATR, cm

-1):

2957, 1709, 1666, 1536, 1459, 1420, 1280, 1186, 1050.

Bis(1,3,9-trimethyl-7-(4-(trifluoromethyl)benzyl)xanthin-8-ylidene)Au(I)

tetrafluoroborate (10)

10b (130 mg, 0.30 mmol), Ag2O (55 mg, 0.24 mmol), distilled acetonitrile (10 mL),

Au(tht)Cl (48 mg, 0.15 mmol) and KI (25 mg, 0.15 mmol). Product: 110 mg (74 % yield).

Crystals suitable for X-ray diffraction were grown by slow diffusion of pentane into a

saturated solution in dichloromethane. 1H NMR (DMSO-D6, 300.13 MHz): 3.22 (s, 3 H, N-

Me), 3.76 (s, 3 H, N-Me), 4.24 (s, 3 H, N-CH3), 5.84 (s, 2 H, N-CH2), 7.45 (d, 3JH-H = 8.1

Hz, 2 H, 2 CHPh), 7.59 (d, 3JH-H = 8.1 Hz, 2 H, 2 CHPh).

13C{

1H} NMR (DMSO-D6, 75.48

MHz): 28.8 (N-CH3), 32.2 (N-CH3), 39.5 (N-CH3), 52.2 (N-CH2), 108.5 (=C-C(O)), 125.9

(m, C-CF3), 128.1 (4 CHPh), 141.1 (CPh-ipso), 141.5 (N-C=), 150.9 (C(O)), 153.4 (C(O)),

187.3 (Ccarbene). 19

F{1H} NMR (DMSO-D6, 282.38 Hz): -61.2 (s, 6 F, CF3), -148.3 (d, BF4

-).

ESI-MS (DMSO-MeOH), positive mode exact mass for [C32H30N8O4F6Au]+ (901.19452):

measured m/z 901.19227 [M-BF4]+. FT-IR (ATR, cm

-1): 2958, 1713, 1668, 1538, 1460,


95

1420, 1322, 1163, 1110, 1053. Anal. Calc. for C32H30AuN8O4F6BF4.CH2Cl2: C: 36.93, H:

3.01, N: 10.44; Found: C: 36.55, H: 3.30, N: 10.87.

X-ray crystallography

Crystals of 3, 4, 7 and 10 were obtained by slow diffusion of pentane through a solution in

dichloromethane. Intensity data were collected on a Bruker Nonius Kappa CCD APEX II at

115 K. The structures were solved by direct methods (SIR92)61

and refined with full-matrix

least-squares methods based on F2 (SHELXL-97)

62 with the aid of the WINGX program.

63

All non-hydrogen atoms were refined with anisotropic thermal parameters. Hydrogen atoms

were included in their calculated positions and refined with a riding model. Crystallographic

data are reported in Tables 2-4. CCDC numbers are respectively: CCDC 951227, 885458,

972032 and 972033.

Table 2: Crystal data and structure refinement for 3.

Empirical formula C9H12AuN4O2I.CH2Cl2

Formula weight 617.02

Temperature 115(2) K

Wavelength 0.71073 Å

Crystal system, space group Triclinic, P -1

Unit cell dimensions a = 7.9219(3) Å α = 79.978(2)°

b = 8.5434(2) Å β = 78.5350(10)°

c = 12.0893(4) Å ϒ = 88.140(2)°

Volume 789.29(4) Å3

Z, Calculated density 2, 2.596 g/cm3

Absorption coefficient 11.619 mm-1

F(000) 568

Crystal size 0.15 x 0.10 x 0.07 mm

2θ range for data collection 2.42 to 27.44 deg.

Limiting indices -10≤h≤10, -11≤k≤7, -15≤l≤15

Reflections collected / unique 9015 / 3563 [R(int) = 0.0442]

Completeness to theta = 27.44 98.8 %

Refinement method Full-matrix least-squares on F2

Data / restraints / parameters 3563 / 0 / 185

Goodness-of-fit on F^2 1.156

Final R indices [I>2σ(I)] R1 = 0.0348, wR2 = 0.0928

R indices (all data) R1 = 0.0350, wR2 = 0.0930

Largest diff. peak and hole 3.310 and -3.693 e.Å-3


96


Empirical formula C18H24AuN8O4.BF4




Crystal system, space group Monoclinic, P21/n

Unit cell dimensions a = 10.7350(4) Å α = 90°

b = 15.2505(5) Å β = 103.438(2)°

c = 14.7786(5) Å ϒ = 90°

Volume 2353.22(14) Å3



F(000) 1360













Empirical formula C30H32AuN8O4.BF4.2(CH2Cl2)




Crystal system, space group Monoclinic, P21/c


b = 18.3831(6) Å β = 94.57°

c = 17.3075(5) Å ϒ = 90°

Volume 3714.75(19) Å3



F(000) 2016


97













Empirical formula C32H30AuF6N8O4.BF4.0.5(CH2Cl2)




Crystal system, space group Orthorhombic, F2dd


b = 29.921(5) Å β = 90°

c = 49.455(5) Å ϒ = 90°

Volume 14857(8) Å3



F(000) 8080











Absolute structure parameter 0.005(12)



98

Cell culture and inhibition of cell growth

The human lung cancer cell line A549 and human ovarian cancer cell lines SKOV3, A2780,

and A2780R (obtained from the European Centre of Cell Cultures ECACC, Salisbury, UK)

were cultured respectively in DMEM (Dulbecco’s Modified Eagle Medium) and RPMI

containing GlutaMaxI supplemented with 10% FBS and 1% penicillin/streptomycin (all from

Invitrogen), at 37°C in a humidified atmosphere of 95 % of air and 5 % CO2 (Heraeus,

Germany). Non-tumoral human embryonic kidney cells HEK-293T were kindly provided by

Dr. Maria Pia Rigobello (CNRS, Padova, Italy) and were cultured in DMEM medium, added

with GlutaMaxI (containing 10 % FBS and 1 % penicillin/streptomycin (all from Invitrogen)

and incubated at 37°C and 5 % CO2. For evaluation of growth inhibition, cells were seeded

in 96-well plates (Costar, Integra Biosciences, Cambridge, MA) at a concentration of 15000

cells/well and grown for 24 h in complete medium. Solutions of the compounds were

prepared by diluting a freshly prepared stock solution (10-2

M in DMSO) of the

corresponding compound in aqueous media (RPMI or DMEM for the A2780 and A2780R or

A549, SKOV3 and HEK-293T, respectively). Afterwards, the intermediate dilutions of the

compounds were added to the wells (200 L) to obtain a final concentration ranging from 0

to 150 M, and the cells were incubated for 72 h. Following 72 h drug exposure, 3-(4,5-

dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) was added to the cells at a

final concentration of 0.5 mg ml-1

incubated for 2 h, then the culture medium was removed

and the violet formazan (artificial chromogenic precipitate of the reduction of tetrazolium

salts by dehydrogenases and reductases) dissolved in DMSO. The optical density of each

well (96-well plates) was quantified three times in tetraplicates at 550 nm using a multi-well

plate reader, and the percentage of surviving cells was calculated from the ratio of

absorbance of treated to untreated cells. The IC50 value was calculated as the concentration

reducing the proliferation of the cells by 50 % and it is presented as a mean (± SE) of at least

three independent experiments.

Preparation of rat Precision-Cut Tissue Slices (PCTS) and toxicity studies ex vivo

Male Wistar rats (Charles River, Kissleg, Germany) of 250-450 g were housed under a 12 h

dark/light cycle at constant humidity and temperature. Animals were permitted free access to

tap water and standard lab chow. All experiments were approved by the committee for care

and use of laboratory animals of the University of Groningen and were performed according

to strict governmental and international guidelines.

PCTS were made as described by de Graaf et al.22

In brief, the intestine tissue was flushed

with ice-cold Krebs-Henseleit buffer. After the removal of the muscle layer, a segment of 10

× 20 mm was embedded in 3 % agarose (low-gelling-temperature agarose type VII, Sigma-

Aldrich, Steinheim, Germany) using a tissue embedding unit. Cores of tissue were prepared


99

from liver and kidney using a coring tool as described. Rat intestine, liver and kidney tissues

were subsequently sliced with a Krumdieck tissue slicer (Alabama R&D, Munford, AL,

USA) in ice-cold Krebs-Henseleit buffer saturated with carbogen (95 % O2 and 5 % CO2).

Intestinal slices (350-450 µm thick and 2-4 mg wet weight) and liver and kidney slices (250

µm thick and 4 mg wet weight) were stored in ice-cold Krebs-Henseleit buffer until

incubation.

PCTS were incubated in 12-well plates (Greiner bio-one GmbH, Frickenhausen,

Austria), at 37°C individually in 1.3 ml Williams’ medium E (WME, Gibco by Life

Technologies, Paisley, UK) with glutamax-1, supplemented with 25 mM D-glucose (Gibco)

and antibiotics (liver: 50 µg/ml gentamicin (Gibco); colon: 50 µg/mL gentamycin + 2.5

ug/mL amphotericin B; kidney: 100 U/ml penicillin G + 100 µg/ml streptomycin)) at pH 7.4

with shaking (90 times/min) in atmosphere of carbogen. Stock solutions of compounds 2 and

4 were prepared in dimethyl sulfoxide at a concentration of 10-2

M (DMSO, VWR,

Fontenay-sous-Bois, France) and stored at 4°C. The final concentration of DMSO during the

PCTS incubation was always below 1 % to exclude DMSO toxicity. Concentration

dependent toxicities of compounds 2 and 4 were evaluated by incubating the human PCTS

with different concentrations of compounds between 0 and 200 µM. Incubation time was 5 h

for intestine and 24 h for both liver and kidney tissues, respectively. After the incubation,

slices were collected for ATP and protein determination, by snap freezing them in 1 ml of

70 % ethanol/x mM EDTA. The viability of PCTS was determined by measuring the ATP

using the ATP Bioluminescence Assay kit CLS II (Roche, Mannheim, Germany) as

described previously.22

The ATP content was corrected by the protein amount of each slice

and expressed as pmol/µg protein. The protein content of the PCTS was determined by the

Bio-Rad DC Protein Assay (Bio-Rad, Munich, Germany) using bovine serum albumin (BSA,

Sigma-Aldrich, Steinheim, Germany) for the calibration curve. ATP data were expressed as

the relative value to the 5 h vehicle control or to the 24 h vehicle control for intestine or

liver/kidney tissue, respectively.

FRET melting assay

Materials and chemicals: All oligonucleotides were purchased from Eurogentec (Belgium)

in OligoGold® purity grade and purified by RP-HPLC at ~200 nmol scale for ds26 and at

~1000 nmol for all other oligonucleotides.

Oligonucleotides: sequences and higher-order structure preparation. The lyophilized

strands were firstly diluted in deionized water (18.2 MΩ.cm resistivity) at 500 µM for

duplex- and quadruplex-DNA constitutive strands and at 1000 µM for all three- and four-

way junction constitutive strands. The actual concentration of stock solutions were

determined, via a dilution to 1 µM theoretical concentration (i.e. 2 µL in 998 µL water for

duplex- and quadruplex-DNA, 1 µL in 999 µL water for three- and four-way junction), via


100

UV-Vis spectra analysis at 260 nm (after 5 min at 90°C), using the molar extinction

coefficient values provided by the manufacturer. All DNA structures were prepared in a

Caco.K buffer, comprised of 10 mM lithium cacodylate buffer (pH 7.2) plus 10 mM KCl/90

mM LiCl. Monomolecular structures were prepared by mixing the constitutive strand (40 µL

at 500 µM) with lithium cacodylate buffer solution (8 µL, 100 mM, pH 7.2), plus KCl/LiCl

solution (8 µL, 100 mM/900 mM) and water (24 µL). Bimolecular structures were prepared

by mixing each of the two constitutive strands

university of groningen gold-based complexes bertrand, benoit · chapter 2 caffeine-based gold(i)...

Documents